Follow up to: The Origin of Truth

In the “The Origin of Truth”, I discussed how knowledge is rooted in the interpretation of previously uninterpreted personal experiences, using a consistent language, which implies imagination and logic. Now I’m going to build on that.

Specifically, how does an uninterpreted personal experience become a piece of knowledge? What makes an interpretation of an experience true or false?

 

What is Truth?

So once we have logic, language, and an imagination, we can start determining true from false. But first, one more thing, what does this “true” even mean?

Let’s start by walking into a room and looking at a chair. Ignoring the whole part about how we know we’re “walking” and how we know we’re in a “room”, let’s focus solely on the “chair” aspect. How do we know that we’re looking at a chair, and not something else?

The answer first comes from language. We know we’re having an experience of a four-legged object with arms that looks designed for sitting, and we also remember that we’ve used an agreed language called English to assign the codeword “chair” to this experience.

But how do we know this is a “true” chair as opposed to a “false” chair?

 

The word “true” itself, like all other words, is also a codeword for a set of experiences. A true chair is very different from a false chair; the adjective “chair” entails additional, specific experiences we will have about that chair. This is kind of like how the adjective “brown” entails the additional, specific experience of the color brown, a visual experience of low frequency photons in low saturation or illumination.

The specific experiences we will have of a true chair is better seen if we know the other words for a false chair — namely “hallucinated chair”, “imagined chair”, “illusory chair”, and “fake chair”. All of these markers of falseness entail certain experiences: the hallucinated chair and imagined chair will entail very few other people seeing the same chair, the illusory chair will perhaps entail trying to sit on the chair and falling through it because it’s a hologram, the fake chair will perhaps entail walking up to the chair and seeing that it was really an elaborate two-dimensional drawing that only looked three-dimensional from where you previously where.

A true chair will entail not just your personal experiences of the chair itself, but experiences of other people agreeing with you that there is a chair there, that the chair supports a certain amount of weight for people sitting on it, that no one will be able to walk through the chair without being obstructed, that the chair will be broken when thrown with enough force, that the chair will fall when dropped from a point of height, etc.

So you can be more and more sure that you see a true chair when you experience those things that indicate the chair is true and when you don’t experience those things that indicate the chair is false.

 

Meaningfully True

What we learned of the chair is capable of being generalized — something is more likely to be “true” the more we have the experiences that indicate truth and the less we have the experiences that indicate falseness. More interestingly and fundamentally, whenever you hold a belief, you should be able to trace that belief back to the experiences that caused you to believe it, and should also be able to imagine experiences that would cause you to stop believing in it. If you can’t, the belief isn’t a belief about anything at all, and is not meaningful and is probably not worth having.

Even the most obvious and basic truths that we hold are grounded in experience. For one example, consider the law of non-contradiction we discussed earlier: we know this law is true because we have always experienced it to be so, and moreover we know that its falseness is an impossibility because we can never imagine it being false, even in principle — and the events we can’t imagine occurring, even in principle, are impossible because that’s how impossible is defined.

 

For second example, consider the statement that “1 + 1 = 2″. Initially, this seems unconnected to experience, because we never experience the essence of “one” alone and by itself, we only experience one as an adjective — one apple, one chair, etc. But the statement “1 + 1 = 2″ is still a statement rooted in experience — it is true that every time we take one object and take another object, we end up with two objects; regardless of what those objects are. Thus we have the generalization “1 + 1 = 2″.

If we work hard, we can even imagine the kind of evidence we would need to see to prove “1 + 1 = 2″ false: we would have to experience taking one object and another object and ending up with something other than two objects, without any objects leaving or any more objects entering. The fact that we cannot imagine this, even in principle, indicates the impossibility of “1 + 1 = 2″ being false.

 

For a third and final example, consider the statement “gravity on Earth is 9.8 meters per second per second”. This is also a generalization that is true of any and all objects capable of falling — that if we set out measuring the time an object takes to fall and the distance it falls, we will always experience these two measurements having the same ratio: for every additional second the object falls unobstructed, the object will cover an additional 9.8 meters for each second before. (On the fifth second, the object will cover 9.8*4 = 39.2 meters.)

This statement itself is also not that difficult to falsify — all we would need to do is measure an object on Earth falling unobstructed (this means correct for air resistance and such) at some other ratio.

 

I can sum this up as follows:

• any statement that is meaningful will entail some experiences must occur and some experiences must not occur
• any statement that is true will entail that (1) all experiences that must occur will occur and (2) no experiences that must not occur will occur
• our degree of confidence in a statement being true is the degree that we actually do experience (1) and (2) being the case.

 

The Powers of Prediction and Goal-Satisfaction

One way we know that the statements we make are very likely to be true is that trusting these statements make predictions about the future that we do indeed experience, and are capable of satisfying our goals.

What if I wanted to build a car? There are many ways I could attempt to build one, such as gluing toothpicks into the shape of a car, or by trying to summon a car by muttering “summon-o car-o” over and over. Only certain ways of building will actually lead me to experience a car — both the toothpick method and the summoning method will consistently lead me to experience failure.

The successful ways of building a car make a prediction of future experience — that I will, in the future, experience a car if I follow these methods. Furthermore, the successful ways of building a car satisfy my goal of having a car — they do not result in the experiences of failure.

And if the reason we want to have knowledge is because it satisfies our goals, this would indicate that the only statements worth assigning as “true” are the ones that actually satisfy our goals — the only “true” methods of car building are the ones that actually build cars. Additionally, this would also indicate that if trusting a certain statement results in us actually experiencing all the things the statement predicted would occur and actually satisfying our goal, then the statement was true.

 

The Powers of Coherent Webs and Coherent Methods

While prediction and goal satisfaction are good ways of testing if individual beliefs are true, there are two additional ways to test if our believes as a whole are true: whether they cohere, or make sense when fit together. To do so, we apply the laws of logic to our beliefs (laws derived from experience) and see if our beliefs fit (an experience of “coherence”).

First is the idea of forming a coherent web: none of the beliefs in our web of beliefs rules out another belief in our web of beliefs. For example, if we have a belief that “apples are always red” and another belief that “some apples are green”, then these two beliefs rule each other out and do not cohere — it is not possible for both beliefs to be simultaneously true as presently defined. Thus, no matter how many successful predictions both beliefs make and how many goals you satisfy with both beliefs, one must be false.

Second is the idea of coherent methods: two methods that should produce the same result will produce the same result, unless we are wrong about our methods or wrong about our results. For example, consider this list of numbers “1 6 2 0 4 3 9 1 8 2 2 7″. We want to know the sum of these numbers. One method is to start at the “1″ and add the numbers, going from left to right. Another method is to start at the “7″ and add the numbers, going from right to left. These two methods should produce the same result, and they do — adding from left-to-right produces 45 and adding from right-to-left produces 45.

If all our beliefs make predictions that turn out to be true, help us satisfy our goals, cohere together logically as a web, and aren’t contradicted by other methods that should get the same results, then we can be reasonably confident that all our beliefs are true.

 

Being Wrong and Postmodernism

Postmodernism typically endorses an idea of “relativism”, or that there is no such thing that is true. …Well, at least, no such things as true besides this sentence.

However, when the postmodernist actually wants results, postmodernism ends up being silly. If there is no such thing as truth, this implies that the summoning method of car design and the engineering method of car design are equally likely to produce a car. However, this is clearly not the case.

We are indeed capable of being wrong: anyone who predicts that some experience will not occur and then finds that it does is wrong. Likewise, anyone who predicts some experience will occur and then finds that it does not is wrong. These people who are wrong may not think they are wrong (or even capable of being wrong), but they will again and again find their goals thwarted as long as they try the summoning method and not the engineering method.

Postmodernism also likes to highlight uncertainty, but how uncertain are we in the superiority of the engineering design over the summoning design? Well perhaps the summoning design works great, it’s just that no one has done it yet. But we don’t have any experiences that would lead us to believe in the efficacy of summoning or even anticipate that summoning could be successful.

Indeed, while we are uncertain in principle that the sun will rise in the sky tomorrow, because there is indeed a chance that it won’t, this is not the same “uncertain” we talk about when we say we are uncertain that the moon is made of cheese. As Eliezer Yudkowsky has said: “it seems worthwhile to distinguish ‘faith’ that the sun will rise in the east just like the last hundred thousand times observed, from ‘faith’ that tomorrow a green goblin will give you a bag of gold doubloons.”

 

Next Up

Now that we have a notion of what makes something “true” or “false” and a notion of what we want out of knowledge, we can work toward creating a model of the world that accomplishes our these goals and gets at “true” knowledge. However, first we have to overcome some philosophical hurdles placed in front of concepts of knowledge, such as the notion of “A Priori” knowledge, questions of whether or not there really is an external world, Cartesian Demons, the possibility of us living in a computer simulation or being brains in a vat, the possibility of everything being a dream, and other issues.

Here’s a sketch of where this series is going and what you can look forward to: The next post will deal with resolving these philosophical hurdles. The post after that will rather thoroughly explain and resolve a lot of critiques about whether this epistemology suffers from “infinite regress” or “circularity”. Then, the post after that will discuss about how we use the logical powers of inference to fully form a model of the world. After all of that, we should be done with epistemology series itself, and can start applying it to knowledge of science and other areas.

Followed up in: The End of Cartesian Demons